BACKGROUND: Several factors influence the pulmonary function values considered normal. In children of short stature, there are difficulties in interpreting the pulmonary function. OBJECTIVE: To assess spirometric values in children and adolescents with short stature and to identify a correction factor to adequately predict the expected values for this population. METHOD: A prospective selection of 77 patients was made, all with short stature and no respiratory disease. These patients were submitted to spirometry, transcutaneous hemoglobin oxygen saturation, chest perimeter measurement, and immediate hypersensitivity testing. Bone age was assessed by wrist X-rays. The data obtained by spirometry (FVC, FEV1, and FEF25-75%) were compared with those of Polgar and Promadhat (1971), predicted in three ways: a) by actual height; b) by height estimated at the 50th percentile for chronological age (CA); c) by height estimated at the 50th percentile for bone age (BA). RESULTS: The mean height was 133.3 ± 13.2 cm, and the deficit in relation to the third percentile was 5.4 ± 6.0 cm. The values obtained for FVC, FEV1, FEF25-75%, were significantly higher than those predicted by actual height. The mean FEV1 obtained was 2.42 ± 0.71 L, and the predicted (actual height) was 2.10 ± 0.64 L; according to the height estimated by BA and CA, the values were 2.27 and 2.86 L, respectively. The mean FVC1 was 2.20 ± 0.6 L, and the predicted was 1.90 ± 0.55 L. With the height estimated for bone age and chronologic age, the predicted values were 2.10 and 2.60 L, respectively. CONCLUSION: Children and adolescents with short stature have higher spirometric values than predicted for their actual height. These findings suggest that the height estimated at the 50th percentile for bone age can be used to evaluate pulmonary function.